Music Mixing Style Transfer: A Contrastive Learning Approach to Disentangle Audio Effects
This addresses the challenge of automating music mixing style transfer for audio producers, though it appears incremental as it builds on existing contrastive learning and source separation methods.
The paper tackles the problem of transferring the mixing style of a multitrack audio to match a reference song, achieving results that closely approximate the reference style and demonstrating robustness in mixture-wise style transfer.
We propose an end-to-end music mixing style transfer system that converts the mixing style of an input multitrack to that of a reference song. This is achieved with an encoder pre-trained with a contrastive objective to extract only audio effects related information from a reference music recording. All our models are trained in a self-supervised manner from an already-processed wet multitrack dataset with an effective data preprocessing method that alleviates the data scarcity of obtaining unprocessed dry data. We analyze the proposed encoder for the disentanglement capability of audio effects and also validate its performance for mixing style transfer through both objective and subjective evaluations. From the results, we show the proposed system not only converts the mixing style of multitrack audio close to a reference but is also robust with mixture-wise style transfer upon using a music source separation model.